Research map / Probabilistic and scientific modeling / Dynamical systems
Resilience Inference for Supply Chains with Hypergraph Neural Network
Zetian Shen, Hongjun Wang, Jiyuan Chen, Xuan Song
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use hypergraph message passing (SC-RIHN) instead of best GNN baseline | supply chain resilience inference · SCR | F1: 0.714 -> 0.770 | Improved |
| use hypergraph message passing instead of GraphSAGE | supply chain resilience inference · TESLA | F1: 0.806 -> 0.856 | Improved |
| add learnable positional embeddings in feature encoder (ablation) | SC-RIHN · SCR | F1: 0.727 -> 0.770 | Improved |
| add learnable positional embeddings in feature encoder (ablation) | SC-RIHN · TESLA | F1: 0.736 -> 0.856 | Improved |
| exclude product nodes from graph-level pooling (ablation) | SC-RIHN · SCR-NR | F1: 0.653 -> 0.709 | Improved |
| use hypergraph reduction (clique, bipartite, firm-only) with GNN baselines (ablation) | ResInf, GIN, GraphSAGE · SCR and TESLA | F1 | Hurt |
| drop product nodes (firm-only reduction) (ablation) | GNN baselines · SCR and TESLA | F1 | Hurt |
| increase observation window T beyond 5 (ablation) | SC-RIHN · SCR | F1 | Hurt |
| increase layers to five (ablation) | SC-RIHN · SCR | F1 | Mixed |
| use only a single-step observation (T=1) (ablation) | SC-RIHN · SCR | F1: about 0.75 | No clear effect |
| random edge removal perturbation | SC-RIHN · SCR-ER | F1: 0.770 -> 0.811 | Improved |
Closest papers by meaning
- Learning Production Functions for Supply Chains with Graph Neural Networks (AAAI 2025 · similarity 0.87)
- Time Series Supplier Allocation via Deep Black-Litterman Model (AAAI 2025 · similarity 0.82)
- Deep Representation Learning for Forecasting Recursive and Multi-Relational Events in Temporal Networks (AAAI 2025 · similarity 0.81)
- LightHGNN: Distilling Hypergraph Neural Networks into MLPs for 100x Faster Inference (ICLR 2024 · similarity 0.81)
- A Cross-View Hierarchical Graph Learning Hypernetwork for Skill Demand-Supply Joint Prediction (AAAI 2024 · similarity 0.81)
- Adaptive Multi-Faceted Service Capabilities Co-Prediction for Nationwide Terminal Stations in Logistics (AAAI 2025 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
